In this section, the author highlights the quality of the research conducted by discussing the reliability and validity of the study and all of its components.
5.1 Reliability
Reliability assesses whether the same study would lead to the same results if conducted by a different researcher or on a different occasion (Saunders et al., 2019). In this case, it is a large concern because of the qualitative research design and the COVID-19 pandemic. To ensure reliability in this study, the author describes in detail the focus of the thesis (4.1), the sample selection (4.2), interview procedure (4.3), interview details and particularities (4.4), and the limitations of this study (7.5), as suggested by Borman et al. (1986). This transparency in the research process is essential to gain reliability.
Furthermore, various interviewee and interviewer biases and errors (Saunders et al., 2019) that occur in this type of research and threaten reliability are addressed:
- The interviewee error affects the ability to respond, thus the performance e.g. bad timing. This type of error was difficult to avoid, in particular due to the COVID-19 pandemic and because it is on the side of the interviewee. However, the researcher took measures to limit it and make sure the respondents felt comfortable conducting the interview already prior to it. In particular, the researcher considered the personal situation in connection with COVID-19 and the psychological impact of it, personal situation, and job or personal responsibilities when selecting the interviewees (see
‘Approval Questions’ in Appendix C). Furthermore, the interviews were scheduled with high flexibility to make sure that both date and time fit the interviewer and interviewee. A time slot of 1:30 hours was also agreed upon to avoid time pressure.
- On the other hand, the interviewer error may lead to a misunderstanding of a response through outside circumstances such as tiredness. Similar to the interviewee error, the pandemic also affects the researcher. Therefore it is difficult to make sure that no error occurred. However, the researcher took all possible precautions to ensure active listening, note-taking and full understanding. Various probing techniques were used (see 4.3) to gain deep insight, understand the true motivations, feelings and thoughts of the interviewees, and elicit the right type of information. To make sure that there were no misunderstandings, the discussions between the interviewer and interviewee
were often summarized and complex answers were re-phrased by the researcher.
Furthermore, the interviews were recorded with the permission of the interviewees, which allowed the researcher to revisit the discussion when analysing the results.
- The interviewee bias might lead to false responses due to answering what the interviewer might want to hear or being scared of answering honestly. To avoid this bias the researcher put emphasis on honest communication before the interview started.
The interviewer also encouraged the interviewee to ask any clarification questions. In general, the interview answers were discussed objectively and the researcher ensured the participants that their opinions will be published anonymized.
- The interviewer bias might lead to a wrong interpretation due to the subjective view of the interviewer. The researcher was fully aware of this possible bias from the beginning and thus tried to counter it by staying objective and trying to understand the different points of view through the probing techniques as mentioned above.
Furthermore, the interview guideline was created in a way to counter this bias due to its open-ended questions. This would ensure that various different answers were given.
Finally, the reader should be aware that one respondent in the sample group is older and thus, does not fit into the same age group as the others. This will be discussed further in section 7.5.
5.2 Validity
Validity is another main concern in this research. We differentiate between three different types: concept validity, internal validity and external validity (Saunders et al., 2012).
Concept validity refers to the research strategy and whether it is able to capture what shall be researched. In this case, whether online semi-structured interviews are the appropriate way to find out about peoples’ perceptions and feelings regarding temporary logo changes. Based on the nature of this research, the author considers the concept validity very high. The researcher has the freedom to explore the topic in detail through semi-structured interviews as well as understand the subjective opinions based on the individual socially constructed meanings.
Secondly, internal validity is about the inferences that researchers make based on the data they elicit. The question that has to be answered here is whether the interpreted results reflect the reality as given by the interviewee or whether it has changed due to misunderstanding or wrong interpretations. To ensure a high level of internal validity the author took the following steps:
1) The author used a predeveloped interview guideline that was kept the same for the first set of the study and adapted in the second set based on the initial findings (see 4.4.4).
Part of this interview guideline were texts and images that were read or shown to the participants, for example: information provided to respondents who were not familiar with the brand, campaign summaries and the corresponding images of the brand logos and the temporary changed logos (see Appendix D). This made it possible for the researcher to compare the individual answers within the two sets.
2) Furthermore, prior to the interview, the author assessed whether there are possible outside factors that could influence the answers given by the respondents, as mentioned above. For more information see interviewee error, interviewee bias and Appendix C.
3) Finally, during the interview, the researcher took steps to ensure full understanding and correct interpretations. For further information see interviewer error.
Finally, external validity/generalisability is different in qualitative than in quantitative research due to its difficulty of applying the results to any other cases than the ones presented.
Therefore, a different type of generalisability – analytical generalisability (Yin, 2003) – is used to translate the results into other real life contexts. Analytical generalisability means reaching a broader theory through generalisation from particulars. In this study, the author tires to reach a high level of it by introducing multiple cases. In that sense, the author sees cases as if they were experiments and in that sense replications, as suggested by Yin (2003). Thus, the author selected cases that maximize generalisability and take various aspects in consideration such as campaign success, campaign goals, industry, etc. (see 4.1) (Kennedy, 1979). Through comparison among one another and determining overarching patterns these cases lead to analytical generalizability and thus broader application (Firestone, 1993). Furthermore, the author uses these unique cases to extending and applying existing theories, while also discussing conflicts within literature. This leads to insightful results with a strong foundation and applicability.